Abstract
Drawing on an empirical study, this article explores the role of immigration detention in Italy by analysing the way a specific rhetoric of ‘dangerousness’ has developed and is being used within the framework of immigration enforcement policies. Our argument is that immigration detention has been transformed into an instrument of crime prevention and ‘social defence’, and that this transformation is fuelled by the central position that the legal categories of ‘risk’ and ‘danger’ have assumed in the regulation of the return procedure. The article contends that immigration law enforcement agencies can make use of immigration detention as a flexible control tool to manage what are perceived as the most problematic populations in urban areas, thus practising a policy of selective enforcement that while not explicitly built along racial and ethnic lines, clearly discriminates among migrants according to their ‘social marginality’ or supposed ‘social dangerousness’.
Introduction
Immigration law and criminal law have somewhat similar functions in defining the boundaries of ‘belonging’. This affinity has been strengthened in recent decades when, in parallel with a progressive tightening of migration controls, we have witnessed what contemporary political and social scientists regard as a process of securitization and criminalization of migration (Guild, 2009; Huysmans, 2006; Palidda, 2013). The complex relationship between migration and criminal justice policies has been explored by many (Aliverti, 2012; Bosworth, 2019; Bosworth et al., 2018; Franko Aas and Bosworth, 2013; Legomsky, 2007; Stumpf, 2006). Situating his argument in the context of a rich historical reconstruction of US deportation law, Daniel Kanstroom (2007) has suggested that the coercive measures provided for by immigration law should not be regarded only as a means of migration control. He argues that, given their impact on migrants’ personal liberty and the use that has been made of them in the course of history, deportation and immigration detention should instead be considered as a flexible means for enacting criminal policies outside the legal framework of criminal law. Kanstroom’s hypothesis, in short, is that deportation law can be used against migrants to achieve objectives that are legally impossible or otherwise more difficult to achieve through the ordinary social control tools provided for by criminal law.
Many other scholars have advanced similar hypotheses by observing that the connections between policing and immigration have expanded as a consequence of the increased blurring of criminal justice and migration governance (Parmar, 2018). The point seems to be that immigration law provides the police with a more easily navigable alternative to the criminal law for dealing with minor offences (Fabini, 2017; Weber, 2012; Weber and Pickering, 2013) or pursuing anti-terrorism objectives (Roach, 2011), and this is precisely the core of our argument here.
Drawing on the results of an empirical study carried out in Italy, in this article we will contend that in a context shaped by reduced capacity within the Italian immigration detention system, pre-removal detention is being used selectively with the main aim of removing from the public sphere some categories of migrants deemed as particularly ‘undesirable’ because of their ‘social marginality’ or supposed ‘social dangerousness’. Our argument is twofold: (1) that immigration detention has been transformed into an instrument of crime prevention and ‘social defence’—aimed at protecting society from dangerous individuals; and (2) that this transformation, fuelled by the central position that the legal categories of ‘risk’ and ‘danger’ have assumed in the regulation of the return procedure, is particularly visible in the practices of Italian immigration law enforcement agencies, all of which have been given wide discretion in defining ‘dangerous mobility’.
In particular, we use the concept of ‘dangerous mobility’ to refer to a complex set of arguments, rhetoric and justifications mobilized by immigration law enforcement agencies responsible for return procedures to justify the adoption of coercive measures affecting migrants’ personal freedom. The research presented in this article shows that the rhetoric of dangerousness is used to make subtle distinctions between ‘reliable’ and ‘dangerous’ migrants from whom ‘society must be defended’ (Foucault, 2003). Our idea is that by defining a specific ‘dangerous mobility’ the police can make use of immigration detention as a flexible control tool to manage what are perceived as the most problematic populations in urban areas, thus practising a policy of selective enforcement that while not explicitly built along racial and ethnic lines, clearly discriminates among migrants according to their ‘social marginality’ or perceived ‘social dangerousness’.
Immigration detention in Italy: An overview
Immigration custody, or detention, has had a troubled history in Italy. The first facilities for the detention of irregular migrants in the process of being removed were opened in 1998, and were called Centri di permanenza temporanea (temporary holding centres). In 2008 their name was changed into Centri di identificazione ed espulsione (identification and removal centres), before being renamed Centri di permanenza per i rimpatri (holding centres for removal) with the enactment of Decree No. 13/2017.
An initial period of growth and consolidation saw the capacity of the detention system in Italy increase from fewer than 1000 beds in 1999 to about 1900 beds in 2011, distributed across 13 detention facilities, while the maximum detention term increased to 18 months. However, this was followed by a period of substantial political disinvestment in immigration detention, apparently prompted by an official report published in 2013 (Ministero dell’Interno, 2013), in which the inefficiency and excessive costs of immigration detention as a tool for return policy were highlighted.
Immigration detention in Italy then declined steadily until 2016. The number of migrants taken into custody fell from 12,112 in 2009, to 2982 in 2016. Similarly, while there were 13 immigration detention facilities operating in 2010, there were just four such facilities operating in 2016, with an overall capacity of 359 beds. This reduction appeared to be a response to an explicit policy plan. Prior to the onset of the so-called ‘refugee crisis’ in which the number of irregular arrivals by sea increased dramatically, public opinion had been shocked by several reports of human rights violations perpetrated inside detention facilities, and by a number of striking protests staged by migrants in response to poor detention conditions. 1 This prompted the centre-left government then in charge to drastically reduce the maximum term of detention to three months (introduced via Law No. 161/2014), and to initiate an overall reassessment of the return policy which subsequently led to a rapid decline in the capacity of the detention system (Cossiri, 2015).
The situation has changed significantly as a result of the so-called ‘refugee crisis’, which saw a marked increase in the number of people landing on Italian shores between 2014 and 2017. In a context in which humanitarian concerns have gradually given way to the need to prevent ‘secondary movements’ of asylum seekers and to implement a more efficient policy for the return of ‘failed’ asylum seekers, the European Commission (2015) has explicitly criticized aspects of Italy’s immigration system. These criticisms have focused on the limited capacity of its detention system and the maximum terms of detention which are well below the threshold set by Directive 2008/115/EC. This ultimately prompted the Government to enact Decree No. 13/2017, which increased the maximum term of detention to six months and envisaged the opening of new detention facilities. The impact of this new policy, further strengthened by the current government with the approval of the Decree No. 113/2018, is yet to be fully realized; in the meantime the capacity of the system has slowly begun to expand and as a result six detention facilities, with a total of 880 beds were operating in June 2018.
Many commentators have already described the poor conditions inside Italian immigration detention facilities (Medici per i Diritti Umani, 2013; Senato della Repubblica, 2016), while sectors of the Italian legal (Di Martino, 2014; Pugiotto, 2014) and social theory (Campesi, 2013; Mazza, 2013) communities have in recent years highlighted the irreconcilability between a form of administrative custody which is exclusively linked to the effective management of migration policies and the most basic constitutional principles. It is not possible here to summarize the extensive international scholarly debate about the legitimacy of immigration detention. It is, however, important to note that case law of national and international supreme courts justified immigration detention as long as it is necessary for the effective enforcement of the return decision (Cornelisse, 2010; Lyon, 2014; Wilsher, 2012).
However, it is precisely in light of this principle of efficiency that immigration detention has been criticized by socio-legal scholars. In particular, Leerkes and Broeders (2010, 2013) have suggested that beyond the official functions assigned to it, immigration detention also performs a number of quite relevant ‘informal’ functions, such as symbolizing state capacity to enforce border controls, deterring uncooperative migrants and incapacitating the most troublesome among them. This can be seen to apply to Italy, where rates of deportation enforcement have always been conspicuously low, and the percentage of migrants effectively removed has always remained below 50% of those taken into custody, while an increasing number of irregular migrants have remained on Italian territory in a state of effective ‘undeportability’ (Fabini, 2019). In these circumstances, it is difficult to understand the actual function that immigration detention plays in the framework of Italian migration policy. In our previous research (Campesi, 2015), we explored the potential of immigration detention in deterring migrants from resisting deportation. We showed that the rationale of deterrence is taken into consideration, especially by those actors involved in the practical management of detention facilities and the implementation of deportation policies. In this article we turn to the work of immigration law enforcement officials who ‘decide to detain’ (Weber and Gelsthorpe, 2000) with the aim of exploring the hypothesis that immigration detention has been assigned a crime prevention function.
Methodology
This article elaborates on our hypothesis by presenting the main results of a wider empirical study of decisions made by Justices of the Peace (JPs), who are responsible in Italy for the judicial oversight of forced return and detention orders (or alternative measures to detention) issued by immigration law enforcement agencies. 2
Hearings are held within 96 hours of the migrant being taken into custody by the police and usually take place within the detention facility. Our study is based on an analysis of 383 return decisions and 281 orders for detention (and alternative measures to detention) included in 426 files relating to judicial hearings for the validation of forced return decisions and orders for detention (and alternative measures to detention) held during the first and fourth quarters of 2015 in Bari and Bologna (see Tables 1 and 2). The inclusion of two study areas offered the possibility of observing the work of two very different JP offices. In the case of Bologna, the detention centre ceased operating in April 2013, while the city of Bari hosts one of the six Italian detention centres that accommodates migrants arrested throughout the country. In Bologna, non-custodial measures prevail, while in Bari detention is the norm.
Types of return decisions issued in Bari and Bologna.
Types of order issued by Questure in Bari and Bologna.
Note: aAll these cases related to migrants kept in detention, although the files did not include a record of the order of detention issued by the Questura.
In addition, we carried out 10 semi-structured interviews with JPs who were in office during the period covered by our empirical analysis. Our study used a mixed methodology, incorporating both quantitative and qualitative analysis of data extracted from case records, although in this article we give less emphasis to our quantitative results. Our goal was to deconstruct the concept of ‘dangerousness’ as it is employed in police and JPs’ practice, and to evaluate the extent to which this concept is mobilized in arguments to justify the use of custodial measures against certain categories of migrants.
‘Risk’ and ‘danger’ as categories of Italian immigration law
The role of social control institutions in governing ‘social dangerousness’ has been the subject of extensive analysis from both social and criminological theoretical perspectives in relation to developments in late modern societies (Harcourt, 2007; O’Malley, 2004; Zedner, 2009). Ashworth and Zedner (2014) are probably among the few who have tried to locate, theoretically and historically, the logic of ‘preventive justice’ shaping contemporary crime control policies. But, in spite of the fact that we follow their argument on the inherently preventive nature of much of the legislative endeavour with respect to migration control, we prefer to refer here to the idea of ‘social defence’, 3 a notion which is deeply rooted in the conceptual traditions of continental criminal law scholarship and criminology (Digneffe, 1998; Fijnaut, 2017) and has profoundly influenced the evolution of the Italian criminal justice system (Garfinkel, 2017; Pires Marques, 2013).
The concept of ‘social defence’ is commonly linked to the transformation of penology at the beginning of the 20th century and to the emergence of indeterminate sentencing and preventive detention—what in continental criminal law are still today defined as ‘security measures’ (Pifferi, 2016). The concept in reality had a much broader scope as it suggested the need to realize a complete change in the aims of the criminal justice system, abandoning traditional reactive models of crime control, based on concepts of retribution and responsibility, in favour of a preventive and proactive approach aimed at ‘defending society’ from potential harms by controlling ‘dangerous’ individuals. 4 In particular, there was an administrative dimension to ‘social defence’ that, at least in Italy, preceded the use of the concept in criminology and penology. This dimension was related to the powers that the Italian police were (and are still) entrusted with to pursue ‘preventive measures’ aimed at neutralizing the danger that certain individuals are believed to present even before they have caused harm to society by committing a criminal offence. 5 The logic of preventive policing has deeply influenced the evolution of Italian immigration law and the related policies of migration control.
In Italy, the management of human mobility has traditionally been associated with the functions the police perform as guarantors of public order and security. Until the enactment of an organic immigration law in 1990, the only rules that regulated foreigners’ entry to and continuing presence in the country were the scant provisions included in the Consolidated Law on Public Security, 6 which essentially regarded foreigners as dangerous subjects to be kept under strict police surveillance. Even when the foundations for the first systematic regulation of immigration were laid down with the enactment of Law No. 39/1990, the ‘public order perspective’ (Caputo, 2006: 173) from which Italy had viewed the rising number of migrants remained essentially unchanged, and the focus was on the regulation of the prerogatives entrusted to the police for the control of irregular migration.
‘Risk’ and ‘danger’ are still crucial legal categories in the provisions regulating the return policy, which are now included in the Consolidated Law on Immigration. 7 For instance, Italian law allows a return decision to be issued against a migrant who is legally staying in the country if an assessment of their possible future behaviour leads immigration law enforcement agencies to regard him or her as ‘dangerous’. This is the case, for example, with return decisions issued by the Minister of the Interior for ‘reasons of public order and national security’, 8 and decisions issued by a criminal court as a ‘security measure’ against ‘dangerous offenders’ according to the definition included in the Italian Criminal Code. 9
The most common situation in which the return decision is justified on account of the migrant’s ‘social dangerousness’ is the one provided for by Article 13(2) of Legislative Decree No. 286/1998, according to which the Prefecture may order the return of those who, ‘according to factual evidence’, could be considered as committed to illicit trades, or are otherwise surviving on the proceeds of criminal activities. This definition of ‘social dangerousness’ is provided for by the Italian Consolidated Law on Public Security, to which immigration law explicitly refers; it is also the legal grounds for the adoption of precautionary measures by the police aimed at neutralizing the danger that certain individuals are believed to present even though they have not been charged with a crime. In this case, immigration law enforcement authorities are granted wide discretion to decide the return of ‘suspect’ or otherwise ‘undesirable’ migrants and to assess ‘social dangerousness’ without any judicial control, unless an appeal against the return decision is filed, explicitly challenging the assessment made by the Prefecture (Savio, 2012: 117).
The legal categories of ‘risk’ and ‘danger’ also play a crucial role in determining how the return decision must be enforced. As a general rule, following the transposition into Italian law of Directive 2008/115/EC, migrants should be granted a term for their voluntary departure, unless a return decision is issued as a consequence of the migrant being considered a ‘dangerous’ or otherwise ‘unreliable’ individual.
According to Italian law, voluntary departure cannot be granted to those whose return is ordered as a result of a criminal conviction, or as a ‘security measure’, or to those whose return has been ordered on account of the need to protect national security or because of the ‘social dangerousness’ of the individual. 10 Voluntary departure also cannot be granted to those whose request for a residence permit has been rejected as groundless or fraudulent, or to those who had already been granted a term for their voluntary departure and had violated the conditions that the police had imposed on them pending their return. 11 In all these circumstances the return decision must be implemented via forced removal, as would be the case when the immigration law enforcement authorities believe that there is the risk that the returnee may abscond.
According to Italian immigration law the ‘risk of absconding’ must be established by considering a number of ‘risk indexes’, such as the possession of a passport, the availability of income and accommodation, and the previous behaviour of the individual concerned. 12 Nevertheless, even when a migrant is granted a term for voluntary departure, this is never unconditional, and the police take a number of steps in order to limit the residual risks. They must, for instance, verify that the individual has adequate income and may impose conditions such as passport surrender; house arrest or forced residence in a given place; or regular reporting to the police. 13
When a forced removal is ordered, and it is not possible to immediately enforce it, the police may adopt a number of measures aimed at preventing the migrant from absconding, and detention is undoubtedly the most coercive option available. Before the transposition into Italian law of Directive 2008/115/EC, pre-removal detention was essentially automatic, and the police could make recourse to it whenever ‘transitory circumstances’ prevented the immediate enforcement of the removal. Nowadays the use of detention must in theory be justified by a ‘risk of absconding’ or by ‘social dangerousness’. In other cases, the police should instead resort to one of the alternatives to detention envisaged by Italian immigration law, including passport surrender; house arrest or forced residence in a given place; or the obligation to report to the police on a regular basis. 14
Evaluation of a migrant’s ‘social reliability’ then comes into play in deciding on the duration of the custodial measures adopted by the police. Before the enactment of Law No. 161/2014, which reduced the maximum term of detention to 90 days, the police could, for example, ask for an extension of detention up to a maximum of 18 months, if the migrant was considered ‘non-cooperative’ or was otherwise trying to hinder the removal. Today, circumstances relating to the personal characteristics or to the behaviour of the individual concerned are no longer taken into consideration when deciding on the duration of detention, except in the case of asylum seekers who are deemed to be a ‘danger’ to public order and security, have already been convicted for serious crimes, or where there is a ‘risk of absconding’, or the asylum application has been filed after the issuing of a return decision with the intention to delay the removal. 15
In conclusion, Directive 2008/115/EC has affirmed the principle that forced removal must be a measure of last resort, to be used only when the migrant does not offer sufficient proof of ‘social reliability’ or is deemed a ‘danger’ to public order and security, 16 and that recourse to detention is legitimate only when—in light of the individual’s risk of absconding or lack of cooperation—less coercive measures cannot be adopted. 17 At the same time Italian immigration law has greatly expanded the discretion attributed to immigration law enforcement agencies in applying these principles in individual cases. This has been done by introducing deliberately generic risk categories (e.g. references to the need to protect national security and to the concept of ‘social dangerousness’ provided for by criminal law), and by building risk indexes that, despite an apparent intention to limit the discretion, cover such a wide variety of circumstances as to leave immigration law enforcement agencies essentially free to adopt the most coercive measures at will. Of course, the adoption of such coercive measures must be validated by a JP but, as the results of our research show, the judicial review process is not particularly effective in limiting the discretion that has been attributed to immigration law enforcement agencies in defining ‘dangerous mobility’.
‘Risk’ and ‘danger’ in police and JPs’ practical reasoning
Around a third of the files examined at the JP offices in Bari and Bologna included return decisions issued on account of migrants’ ‘dangerousness’ (see Table 1). They included 135 return decisions issued by Prefectures because the individuals concerned were judged as ‘socially dangerous’ according to police law, as well as 16 return decisions issued by a criminal court as ‘security measures’ against habitual offenders, and one return decision issued directly by the Ministry of Interior for reasons of ‘national security’. In this section, we analyse this sample in detail to understand how immigration law enforcement agencies, and especially Prefectures, evaluate migrants’ ‘social dangerousness’.
Our analysis (see Table 3) found that in 30% of these 135 cases Prefectures explained their assessment of ‘social dangerousness’ by merely paraphrasing the text of the law and including a list of crimes allegedly committed by the migrant concerned. Reference to specific criminal convictions is made in 50% of cases, while in a number of cases (18 cases; 13%) reference is made only to what in common parlance are referred to as ‘police records’; that is, the fact that an individual has been reported as a suspect. Only in a minority of cases (23% of the total) did Prefectures include a thorough evaluation of a migrant’s ‘criminal profile’ in their return decision.
Evaluation of migrants’ ‘social dangerousness’ by Prefectures.
Even more standardized is the assessment of the ‘risk of absconding’ that the Prefectures undertook in order to decide whether or not to grant a term for voluntary departure. Our analysis shows that in most cases the return decisions included a list (57% of cases) or a multiple-choice sheet (10% of cases) replicating the risk indexes provided for by the law (Table 4). Only 22% of return decisions included a more detailed assessment of the ‘risk of absconding’, but even this more thorough qualitative analysis reveals repetitions in the language used by Prefectures, with standardized phrases recurring almost word for word in several return decisions.
Approach to evaluating migrants’ ‘risk of absconding’ by Prefectures.
To give an idea of how the assessment of the ‘risk of absconding’ is carried out, we have reproduced the most commonly recurring formulation of words in the return decisions issued by the Prefectures in our sample:
Given that Mr […] is believed to be at risk of absconding, i.e. there is a danger that he may try to evade return if he is granted a term for his voluntary departure, on account of the fact that:
- he lacks a valid travel document; - he does not appear to have any concrete interest in returning to his home country; - he has not provided, nor is he able to provide, any financial guarantees from legitimate sources of income useful for this purpose [return]; - he lacks a stable residence or domicile where he can be easily contacted by the police; - he lacks stable employment and has shown no social integration in Italy. (Prefecture of Milan) Considering that, in light of the evidence, and also taking into account the principle of proportionality of the measures to be adopted in order to ensure the enforcement of return decisions, the forced removal of Mr […] from state territory must be ordered given that there is a risk of absconding as provided for by Article 13(4bis), Legislative Decree No. 286/1998, since the foreigner:
- has not complied with previous orders to leave the country issued by the competent authority, pursuant to Article 13(5) and 13(13), and Article 14, Legislative Decree No. 286/1998; - lacks a passport, or any other valid travel document; - lacks any documentation attesting to the availability of accommodation where he can be easily contacted by the police, and is not in a position to provide financial guarantees demonstrating legitimate income; - appears to have previously provided false information or falsely attested his personal identity. (Prefecture of Padova) Considering that Mr […] is to be considered ‘at risk of absconding’ pursuant to Article 13(4bis), Legislative Decree No. 286/1998, i.e. there is a danger that he may evade the return decision if he is granted a term for his voluntary departure given that:
- he has declared that he does not want to return to his country of origin; - he is not in possession of a valid travel document; - he lacks a stable residence or domicile where he can be easily contacted by the police; - he has not provided, nor is he able to provide, any financial guarantees from legitimate sources of income useful for this purpose [return]; - he has not complied with a previous order to leave the country issued by the competent authority, pursuant to Article 13(5) and (13), and Article 14, Legislative Decree No. 286/1998. (Prefecture of Bari)
Similar standard wording also appears in the orders for detention (and alternative measures to detention) issued by the police following a forced removal order—although it should be noted that in 34% of the files we examined there was no record of the orders issued (see Table 2). By analysing the Questura orders that were included in the JP files, we found that in 56% of cases a measure restrictive of a migrant’s personal freedom (be it detention or another non-custodial measure) had been requested by the police on account of a ‘risk of absconding’, assessed using the same list of indexes (53%) or multiple-choice sheet (2.8%) commonly employed by the Prefectures. In 29% of cases the Questura simply referred to the evaluation provided by the Prefecture in their return decision (Table 5).
Evaluation of migrants’ ‘risk of absconding’ by Questure.
The practice of including in return decisions and orders for detention (and alternative measures to detention) a list of circumstances that the Prefectures and the police simply ‘tick’ allows immigration law enforcement agencies to presume there is a ‘risk of absconding’, thus betraying the spirit of the standards included in Directive 2008/115/EC. According to EU rules, 18 the need to resort to coercive measures such as forced removal and detention should be assessed in each specific case, taking into account the actual circumstances of the individual concerned, and should not be based on abstract risk indexes defined in advance. It is no coincidence that the assessment of the ‘risk of absconding’ made by the police is hardly ever questioned by JPs during judicial hearings. The grid of pre-printed risk indexes becomes a truly ‘epistemological cage’ forcing JPs’ reasoning into a set of predefined parameters. The data collected during our research show a clear tendency among JPs to accept the assessment of the supposed ‘risk’ or ‘danger’ posed by the individual presented by immigration law enforcement agencies. This is undoubtedly due to the persuasive nature of the risk indexes, which constrain JPs’ reasoning even in the face of detention orders that are ‘poorly justified’ or clearly ‘pre-printed’ (JP, Bari).
In the rare cases where JPs do question the assessments made by the immigration law enforcement agencies, this is usually because the migrant’s lawyer has gone beyond a merely formal opposition to explicitly challenge the assumptions on which removal decisions and detention orders are grounded. However, the data collected during our research show an overall weakness in the defensive strategies used by legal representatives: in 29% of JP hearings held in Bari, and 39% of such hearings held in Bologna, lawyers did not present any arguments in support of their clients. However, despite the valuable efforts of some lawyers, the grid of presumptive circumstances used by immigration law enforcement agencies cannot be seriously challenged in the limited time allocated for each hearing held before a JP. As two JPs explained,
They [the lawyers] bring you documents that they received by fax, but what matters is bringing some real evidence. For example, many say they have a child with an Italian partner, but I need to see some official document attesting the paternity, otherwise how can I set them free in these circumstances? (JP, Bari) Well, the risk of absconding… Yes, because often during the hearings lawyers say: ‘Judge: there is no risk of absconding here, he has a partner who is pregnant, he has a home’. Then the alleged partner makes a statement, but for me this is not enough… I mean, how can I from a sheet of paper…? Given the situation in which I have to decide, I’m not in a position to hear witnesses. How can I take for granted a declaration that I receive by fax? (JP, Bari)
To conclude, it is the practice of immigration law enforcement agencies to define ‘risk’ and ‘danger’ using highly standardized formulas based on a grid of risk indexes, which, in turn, largely shape the reasoning of JPs. The fact that JPs, in the main, accept the case presented by immigration law enforcement agencies is generally justified by the ‘lack of time to make distinctions [between cases]’ (JP, Bari). As the above quotations suggest, lawyers need to make considerable efforts in order to overturn the arguments on migrants’ ‘social dangerousness’ and ‘risk of absconding’ included in removal decisions and detention orders.
Defining ‘dangerous mobility’
The standardization of the assessment of migrants’ ‘social dangerousness’ and ‘risk of absconding’ is certainly an indication of the fact that immigration law enforcement agencies do not need to present complex arguments in order to convince the JPs of the legitimacy of the restrictive measures taken. JPs are essentially concerned about the formal correctness of the administrative action. There is a shared knowledge between immigration law enforcement agencies and JPs which is hardly questioned during judicial hearings, and this knowledge is reflected in the repetitiveness of the discursive strategies used by both parties. By analysing these discursive strategies, we can derive valuable insights into how the social figure of the migrant, whose personal freedom might be restricted, is constructed in practice. However, this social figure is rather complex, and has at least three distinct dimensions. A first dimension is linked to the migrant’s supposed ‘social dangerousness’, usually inferred from previous contacts with law enforcement agencies. A second dimension is linked to their alleged ‘social reliability’, which is essentially related to the individual’s previous migration path. Finally, a third dimension is linked to their ‘social marginality’, which is associated with the migrant’s perceived degree of socio-economic integration in the destination country.
A ‘dangerous’ migrant is usually the one whose return is sought on account of their criminal behaviour or previous contacts with law enforcement agencies, even if a clear reference to specific criminal convictions is often missing. In some cases, return decisions include a list of police records, accompanied by a more or less colourful paraphrasing of the concept of ‘social dangerousness’ as provided for by the Consolidated Law on Public Security (see Table 3). Our more in-depth analysis of the crimes listed for the 50% of cases in which a specific reference to a criminal conviction was made shows that, alongside the 27% of migrants who are considered ‘dangerous’ for having been convicted or even simply reported for a crime involving violence against people or property, there is a further 10% who have been convicted or simply reported for offences under immigration laws, or otherwise related to resisting or circumventing migration controls. In 13% of cases migrants had indeed been reported only for offences such as resisting arrest, false attestation of identity or refusal to be identified. A significant proportion of those considered to be ‘dangerous’ are in essence burdened by criminal convictions or police records which are merely a reflection of the criminalization of irregular migration and of the intense police surveillance that migrants are subjected to.
When the return decision contains a full description of the supposed ‘criminal profile’ of the migrant concerned, very often this is based on a number of stereotypical factors emphasizing the alleged ‘social marginality’ of someone who ‘lives on expedients, or on the proceeds of criminal activities’ (Prefecture of Aquila), has not showed ‘any will to integrate’ (Prefecture of Modena) or has been repeatedly stopped by the police together with ‘other clandestine immigrants usually engaged in illegal trafficking’ (Prefecture of Perugia). But sometimes it may also be suggested that the migrant has shown a clear ‘propensity for crime’ (Prefecture of Perugia), a ‘nature prone to crime’ (Prefecture of Terni) or a ‘strong tendency to commit crimes’ (Prefecture of Reggio Calabria). However, these statements are never the outcome of an in-depth analysis of a migrant’s previous criminal conduct. The particular gravity of the offence or the circumstances surrounding its commission are rarely taken into consideration, and immigration law enforcement agencies are simply satisfied with citing previous ‘criminal’ or ‘police’ records.
In one of the few cases in which the defence went so far as to challenge the elements on which the assessment of the migrant’s ‘social dangerousness’ was grounded, the return decision included the usual list of previous records for theft, aggravated theft, drunkenness and drug dealing, adding that the migrant ‘has never had work’ and ‘has always frequented ex-offenders and drug addicts’ (Prefecture of Bologna). However, during the subsequent hearing before the JP, the migrant’s lawyer presented evidence contradicting these claims, thus debunking his alleged ‘criminal profile’. This eventually convinced the JP not to validate the forced removal order issued by the Prefecture of Bologna.
For the most part, the crimes allegedly committed by those considered ‘dangerous’ relate to their subsistence. The most frequently cited are offences such as drug dealing, theft, stealing and robbery. On the one hand, this is somewhat understandable, as those most likely to be detained come from the most marginalized groups, who often survive on street criminal economies. On the other hand, it is interesting to note that many return decisions emphasize the ‘social alarm’ provoked by street crime. This is a rhetorical strategy that the Prefectures seem to employ as a way of further supporting their statements on the ‘social dangerousness’ of individuals whose criminal histories boast, at most, some police records for dealing small amounts of drugs or other petty offences. In particular, drug-related offences are considered an index of ‘social dangerousness’ because they imply, in the judgement of Prefectures, that the migrant belongs to a ‘criminal organization’ (Prefecture of Perugia), or are considered as evidence of his/her ability to move ‘with some confidence’ within a ‘criminal context’ (Prefecture of Terni), even though the individuals concerned have generally not been previously convicted of being a member of a criminal organization. 19
Many of the migrants concerned have, however, already violated the terms for their voluntary departure or the entry ban that usually follows the issuing of a return decision or have in the past provided a false identity and do not have valid travel documents (see Tables 6 and 7). These are circumstances that usually justify forced removal or the issuing of an order for detention (and alternative measure to detention) and may be considered as crucial elements of the ideal-typical profile of the ‘unreliable’ migrant at ‘risk of absconding’. A migrant’s ‘unreliability’ is clearly deduced from any previous attempts at circumventing controls or hindering the removal procedure; however, it is interesting to note that one of the factors that seems to have the greatest impact on the assessment of the ‘risk of absconding’ is the so-called ‘absence of will to return to the country of origin’.
Main circumstances considered by Prefectures in evaluating migrants’ ‘risk of absconding’. a
Note: aPercentages are calculated based on all the 357 return orders issued by Prefectures, excluding 17 orders where no information on evaluation was available and 22 orders where the risk of absconding was stated by mere reference to articles 13(4) and 13(4bis) of Legislative Decree No. 286/1998 (see Table 4).
Main circumstances considered by Questure in evaluating migrants’ ‘risk of absconding’. a
Note: aPercentages are calculated based on 153 orders for detention (and alternative measures to detention) where the risk of absconding is assessed using a list of indexes, a multiple-choice sheet or a more detailed discursive assessment.
How immigration law enforcement agencies infer migrants’ interest (or absence thereof) in returning to their country of origin is never entirely clear; though it seems that, rather than looking at previous behaviour, agencies attach particular importance to the statements made by migrants at the time of their arrest. However, in the opinion of some JPs, this can lead to paradoxical results:
The ‘risk of absconding’ is considered to be present every time a foreigner declares that he/she does not want to return to his country of origin. As soon as he expresses his/her intention not to return spontaneously to his country of origin, immediately we consider him/her to be ‘at risk of absconding’. In reality, what really happens here? What happens is that if someone is prepared and tells me: ‘I want to go back to my country’, we deduce that there is no ‘risk of absconding’. If he/she speaks naturally, as an honest person, and says: ‘No, I would like to stay in here Italy to work’, we believe that he/she is at ‘risk of absconding’. This is absurd! Because in reality it [the risk of absconding] is assessed on the statements made by the foreigner and the absurd thing is that the more natural and honest the foreigner is, the more we assume there are grounds to forcibly repatriate him/her. (JP, Bari)
However, cases in which immigration law enforcement agencies make their assessment of the ‘risk of absconding’ solely on the basis of what are usually considered indexes of a migrant’s ‘social reliability’ are not common. In fact, these indicators are more often than not combined with other indexes related to a migrant’s ‘marginality’.
The figure of the ‘marginal’ migrant has fairly clear traits and is characterized by a lack of accommodation and stable employment, as well as by an inability to provide adequate financial guarantees regarding their eventual voluntary departure. These circumstances are included in a significant proportion of the removal decisions and orders for detention (and alternative measures to detention) we analysed (see Tables 6 and 7). In the vast majority of assessments of the ‘risk of absconding’ immigration law enforcement agencies combine aspects of different indexes, thus producing a sort of hybrid figure, combining elements of the ‘unreliable’ and the ‘marginal’. Ultimately, migrants are ‘unreliable’ because they lack employment and a stable residence where the police can contact them, and ‘marginal’ because the lack of personal documents prevents them accessing regular employment or stable accommodation. This is explained by a JP in the following terms:
But the Prefectures tell us, or rather the law gives us a clear steer that the ‘risk of absconding’ can be inferred from the fact that the foreigner does not have a job, does not have a home. He does not have a fixed abode… But how can he have a fixed abode if he is ‘clandestine’? [sic] Given his position he cannot pay regular rent. So, it’s a bit of a funnel, let’s say, a bottleneck produced by the law and by the interpretation of the law offered by Prefectures in order to justify the adoption of coercive measures. There have been so many foreigners who [during the hearings] say that… When we have to decide on potential alternatives to detention, such as reporting obligations, where you clearly need a residence… They tell me: ‘I live with four fellow countrymen in a place where we do not pay regular rent…’ Of course, you don’t. If you’re a ‘clandestine’ [sic] how can you have a regular rent? (JP, Bari)
Although the categories of ‘dangerousness’, ‘unreliability’ or ‘social marginality’ are distinct, in practice the overall criminalizing effect of migrants repeatedly attempting to escape migration controls produces an overlap, as perceived by JPs. However, the dimension of ‘dangerousness’ ultimately prevails. During their stay in Italy, many of the migrants in our study had already attempted to resist police controls, and during the hearings before the JPs, they often acted intemperately or openly resisted the removal procedure. This behaviour is then perceived as further indication of the fact that immigration law enforcement agencies are dealing with ‘really dangerous migrants, people with whom you should be careful’ (JP, Bologna).
If we accept that the rhetoric on migrants’ ‘dangerousness’ is a crucial component of the deportation machine and that it is built on assumptions that immigration law enforcement agencies and JPs essentially share, then we can understand why any attempt to resist removal may be regarded as conclusive evidence of a migrant’s ‘social dangerousness’. Once again, we are faced with a self-reinforcing rhetoric. Migrants are not ‘dangerous’ because they are ‘irregular’, but they are ‘irregular’ because they are inherently ‘dangerous’; that is, they deliberately exploit their position of irregularity, and perhaps even seek it, in order to try to evade controls and survive in the shadows.
Closing remarks
Going back to our research question regarding the specific role that detention plays in the framework of migration control policies, we conclude that, while the use of immigration detention in Italy decreased between 2013 and 2016, it was also deployed more and more selectively. According to two circular orders issued by the Ministry of Interior, 20 when the police ask for a migrant to be assigned a place in an immigration detention facility any factors—such as previous criminal convictions or ‘police records’—relevant to the assessment of their ‘social dangerousness’ must be specified. This is clearly reflected in the result of our research. In Bologna, where non-custodial measures prevail, in very few instances are migrants subject to a return decision because they are considered ‘dangerous’ and the evaluation of their ‘risk of absconding’ focuses on the dimension of migrants’ ‘unreliability’ and ‘marginality’. In contrast, the alleged ‘social dangerousness’ of the migrant concerned was the main justification for the return decision in more than the 40% of the cases considered by our research in Bari, and their custody was routinely justified on account of the need to isolate those who are considered a ‘danger’ to the community.
While it should not be inherently surprising that the police and JPs, who are in Italy responsible for the removal procedure, make use of the rhetoric of ‘dangerousness’ as a justification for detaining migrants (given that it was the so-called Return Directive that established that detention should be considered a measure of last resort for migrants ‘at risk of absconding’), we believe that they abuse the discretion bestowed by the Italian immigration law in a way that is altering the very nature of immigration detention and its functions. Originally intended as a precautionary measure of last resort to be used to ensure the effective enforcement of removal decisions, immigration detention largely fails to comply with its official function. Given that less than a half of those taken into custody are actually removed, it could be argued that immigration detention in Italy has become a surrogate crime prevention measure used by the police for the purpose of managing allegedly ‘dangerous’ populations in urban areas. It has thus become a policing tool that is useful not so much at the state level for deporting irregular migrants, but at the local level for banishing them from public spaces. 21
Our research suggests that in Italy immigration detention may be used as a flexible means of crime prevention outside the legal framework of the criminal law, thus supporting many of the assumptions of the scholarly literature that has explored the complex relationship between migration and criminal justice policies. In particular, some criminologists in the emerging field of ‘border criminology’ (Bosworth, 2019; Bosworth et al., 2018; Turnbull, 2017), have argued that the increasing relevance of immigration detention may be read as an example of the expansion of penal logic and practices into migration control policies. Although we consider this idea to be convincing at a general level, we believe that the results of our research invite a degree of theoretical caution when looking at immigration detention through the lens of the concept of ‘punishment’.
Our argument is that immigration detention should essentially be regarded, not as a form of punishment as strictly understood. Instead, it should be seen as an example of the increasing influence of the logic of ‘security governance’ and ‘preventive control’ (Ashworth and Zedner, 2014; Zedner, 2009), according to which law enforcement agencies are afforded expedited control tools which operate at the margins of the criminal justice system and aim at maximizing their capacity to anticipate alleged threats and contain risks. Under contemporary migration laws and policies, the exercise of coercive powers is not justified on account of legal concepts such as ‘individual responsibility’ and ‘culpability’, typically associated with the logic of ‘punishment’, but rather is based on the construction of abstract typologies of ‘dangerous individuals’ perceived as presenting risks to society. What our empirical research has shown is that migrants placed in detention are never deprived of their liberty on account of harm done, but because they are deemed somewhat irresponsible and untrustworthy.
Another point of interest is that the rhetoric employed by the Italian law enforcement agencies recalls the rhetoric of criminological positivism. This is why we refer to the concept of ‘social defence’, an idea which is probably less common in the English-speaking context, where concepts such as ‘prevention’ and ‘incapacitation’ are preferred. The idea of ‘social defence’ is taken from the work of Enrico Ferri (1900). He contrasted ‘social defence’ with the ideas of ‘punishment’ and ‘individual responsibility’, concepts perceived as typical of what he dubbed the ‘classic school’ of penology, with the intent of redesigning the criminal justice system and transforming it into an instrument for the neutralization of ‘dangerous individuals’. As we have seen, law enforcement agencies speak the language of ‘dangerousness’ and ‘criminal tendency’ by using concepts that, in spite of being somewhat scientifically discredited, have survived in the terminology used in Italian police laws. These discursive formations (Foucault, 1972) are used to stigmatize and exercise control over those who are perceived as ‘dangerous’ mobile individuals, who are allegedly always ready to evade controls.
However, in spite of the fact that social defence theory was originally strongly intertwined with a racial perspective on crime and deviance which has been crucial in the production of Italian national identity (Teti, 1993), the rhetoric employed by the Italian immigration law enforcement agencies makes no explicit reference to migrants’ nationality or ethnic origin. While a number of studies have already shown how migrant criminalization in contemporary Italy is supported by an increasingly racialized public discourse (Dal Lago, 1999; Palidda, 2009, 2013), our research seems to suggest that immigration law enforcement agencies are somewhat able to neutralize ‘race’ in their reasoning on migrants’ ‘dangerousness’. Even in cases where removal orders emphasize the ‘social alarm’ caused by certain forms of crime, in which ethnic stereotypes seem to resurface, no explicit argument based on the nationality or ethnic group is ever presented. This probably suggests that the question of the relationship between nationality, race and the selective control of irregular migration in Italy is worthy of further research aimed at exploring the role legal categories of ‘risk’ and ‘danger’ may have in producing an apparently ‘racially neutral form’ of migrant profiling.
Footnotes
Acknowledgements
The article has been jointly conceived and written by the two authors. According to Italian academic folkways, the authors acknowledge that Giuseppe Campesi has written the odd pages and Giulia Fabini the even. The authors would also like to thank the two anonymous reviewers and the editor of this special issue for their comments on the draft version of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Most of the research for this article has been carried out within the framework of a research project sponsored by Open Society Foundation (PI Enrica Rigo, University of Roma Tre).
